Sex-Specific Differences in Serum Kallikrein-8 (KLK8): An Exploratory Study

Abstract

Background:

There are indications for sex-specific differences regarding the association between kallikrein-8 (KLK8) and cognitive impairment in early stages of Alzheimer’s disease for which KLK8 may be an early blood-based biomarker. These may be due to different levels of sex hormones. To correctly interpret KLK8 blood concentrations, sex-specific analyses are needed.

Objective:

The aim of our exploratory study was to investigate sex-specific differences in blood-based KLK8 in participants of the population-based Heinz Nixdorf Recall study with different cognitive status and the association between KLK8 and sex hormones.

Methods:

In 290 participants (45% women, 69.7±7.4 years (mean±SD)) we investigated sex-specific serum KLK8 differences between cognitively unimpaired (CU, 43%) and cognitively impaired (CI) participants and the association between KLK8 and dehydroepiandrosteronsulfate (DHEAS), estradiol and testosterone, using adjusted multiple linear regression.

Results:

The mean±SD KLK8 was similar for CU men (808.1±729.6 pg/ml) and women (795.9±577.7 pg/ml); adjusted mean-difference [95%-CI]: –95.3 [–324.1;133.5] pg/ml. KLK8 was lower in CI women (783.5±498.7 pg/ml) than men (1048.4±829 pg/ml); –261 [–493.1; –29] pg/ml. In men but not women, there was a weak indication for a positive slope between estradiol (11.9 [–0.4;24.3] pg/ml) and DHEAS (1.4 [–0.5;3.3] pg/ml) with KLK8, while testosterone had no impact.

Conclusions:

The results suggested a different role for KLK8 in the development of cognitive impairment in men and women, potentially influenced by sex hormones. To use blood KLK8 as an early biomarker, further research on hormonal regulation of KLK8 expression is needed as a part of the investigation of the KLK8 involvement in cognitive impairment and Alzheimer's disease pathology.

Keywords

Alzheimer’s disease amnestic mild cognitive impairment DHEAS estradiol Heinz Nixdorf Recall study kallikrein-8 KLK8 non-amnestic mild cognitive impairment subjective cognitive decline testosterone

INTRODUCTION

The serine protease kallikrein-8 (KLK8) is expressed in many different body tissues and secreted into the blood circulation [1, 2]. It is highly expressed in the central nervous system, especially in the hippocampus, amygdala and regions of the limbic system, which are known to regulate learning processes and memory functions [1]. Recent studies suggest an association between the overexpression of KLK8 and cognitive impairment in early stages of Alzheimer's disease (AD), for which KLK8 may be an early blood-based biomarker [1 –3]. Elevated concentrations of KLK8 were shown in AD patients, compared to cognitively healthy individuals [2] and in participants of a population-based study with the AD precursor amnestic mild cognitive impairment (MCI) [4]. Also participants of a population-based study with subjective cognitive decline (SCD) (which can be a precursor symptom of MCI or dementia [5]) had higher average blood KLK8 values compared to cognitively unimpaired [6]. In contrary, participants of a population-based study with non-amnestic MCI (which can be a precursor of non-AD dementia forms [7]) did not have higher average blood KLK8 values [8]. Thus, blood-based KLK8 could be a very early biomarker to diagnose AD even in preclinical stages.

In previous studies, sex differences in the expression of KLK8 were noticed. Keyvani et al. reported that hippocampal KLK8 was not only higher in AD patients compared to healthy individuals post mortal, but also in the brains of AD affected and healthy women compared to men [9]. While Teuber-Hanselmann et al. and Goldhardt et al. did not find any sex differences of KLK8 in the blood or cerebrospinal fluid [2, 10], Herring et al. reported that KLK8 was higher in female transgenic mice compared to males [3]. Some positive effects on the AD pathology generated by KLK8-knockdown in mouse models such as an elevation of spine density and the structural plasticity marker fibronectin, the upregulation of Aβ efflux transporter lipoprotein receptor-related protein 1 or the downregulation of Aβ influx transporter RAGE (receptor for advanced glycosylation end product) could only be detected in female and not male mice [3]. AD is known to have higher prevalence and severity in women [9, 11] and AD pathology is suspected to be different in sexes [12]. KLK8 is suspected to show off sex-specific differences [3, 9] and questions about the expression, the regulation, and the role of KLK8 in AD in men and women arise.

Steroid hormones appear to play a role in the regulation of the KLK8 expression [13, 14], but the current state of research is inconclusive. There is evidence for an increasing effect on KLK8, especially by estradiol, but also by testosterone and other steroid hormones in cell cultures [13 –15]. Testosterone has been demonstrated to increase KLK8 in vivo [16], while other cell culture experiments have shown positive effects on AD pathology (such as an improved microglial viability and Aβ-clearance) by testosterone administration [9] and a genomic study showed an association between higher testosterone levels and a lower AD risk [17]. Previous experiments have shown that estradiol but not testosterone is capable to induce KLK8 production in neuronal and microglial cell lines, suggesting that estrogen-induced KLK8-overproduction is responsible for higher KLK8 levels and more pronounced features of AD in females [9], while present literature suspects testosterone is more likely to have a protective role against AD [9 , 19]. However, sex hormones themselves are thought to have neuroprotective functions and to be associated with a lower AD risk [20, 21], which may be contradictory to the previous theories.

Nothing is known about sex differences in KLK8 in a population-based setting. The aim of our exploratory study was to investigate sex-specific differences in blood-based KLK8 in participants of the population-based Heinz Nixdorf Recall study with different cognitive status and the association between KLK8 and sex hormones.

METHODS

Study population

The study population consisted of selected participants of the population-based Heinz Nixdorf Recall study (Recall = Risk Factors, Evaluation of Coronary Calcification and Lifestyle) at the University Hospital Essen in Germany. The study design has been described previously [22]. In brief, the participants were 4814 randomly selected inhabitants of the Ruhr area living in Bochum, Essen and Mülheim/Ruhr (Germany), aged 45 to 75 years [23]. After the baseline examination (T0) in 2000–2003, there were further examinations after 5-years (T1) and 10-years (T2) and participants still receive annual questionnaires. The data analyzed in this exploratory study was originally generated as a part of investigations of the association between KLK8 and SCD [6], amnestic MCI [4], and non-amnestic MCI [8], the results are already published. Because of the data generation for those previous investigations we had the following exclusion criteria: objective cognitive decline at T1 or T2 (according to the results of a standardized cognitive assessment performed at T1 and extended at T2, details were published previously [24, 25]); missing or incomplete information about the cognitive status, or dementia (medical diagnosis according to the DSM-IV criteria [26] or self-reported intake of cholinesterase inhibitors (code: N06DA) or anti-dementia drugs (code: N06DX) [27]); the presence of cancer or stroke before T2; Parkinson's disease; inflammatory diseases (Crohn's disease, ulcerative colitis, rheumatoid arthritis, ankylosing spondylitis); a high sensitive C-reactive protein (hsCRP) >1 mg/dl; invalid KLK8 measurement. Therefore, our final analysis population consisted of 290 participants: 37 participants with amnestic MCI, 55 with non-amnestic MCI, 73 with SCD, and 125 cognitively unimpaired (CU). The flow-chart of the study population is provided in Fig. 1. The study was approved by the Institutional Review Board of the University of Duisburg-Essen and followed the established guidelines of good epidemiological practice. All participants obtained written informed consent.

Fig. 1

Flowchart of the study population. ADL, Activities of daily living; aMCI, amnestic mild cognitive impairment; CU, cognitively unimpaired; hsCRP, high-sensitive C-reactive protein; KLK8, kallikrein-8; naMCI, non-amnestic mild cognitive impairment; SCD, subjective cognitive decline; T1, first follow-up visit; T2, second follow-up visit.

Cognitive status

Because of the original aim for the data generation we had data of 290 participants with four different cognitive groups according to the results of the standardized cognitive assessment performed at T1 and T2: CU (cognitively unimpaired at T1 and T2), SCD (CU at T1 and subjective cognitive decline at T2), amnestic MCI and non-amnestic MCI (CU or SCD at T1 and amnestic/non-amnestic MCI at T2 according to the criteria of Winblad et al. [28]). Details of neuropsychological assessment for cognitive diagnoses have been described elsewhere [24, 25]. Participants with SCD, amnestic MCI and non-amnestic MCI where also pooled into the group cognitively impaired (CI).

Measurement of KLK8

The measurement of the blood-based KLK8 was performed in the laboratory of the Institute of Neuropathology of the University of Duisburg-Essen using ELISA (enzyme-linked immunosorbent assay; #EK0819, Boster Biological Technology, Pleasanton, California, USA). Serum samples were taken at T2, frozen at –80∘C for 6.9 to 9.9 years, diluted 1 : 2 in sample buffer. The three experienced experimenters were blinded to the participant's cognitive status and followed the manufacturer's instructions. The serum KLK8 concentration was measured in duplicate and is given in pg/ml.

Measurement of sex hormones

The serum concentrations of the sex hormones DHEAS (dehydroepiandrosteronsulfate), estradiol and testosterone were measured at T2 in the Central Laboratory of the University clinic of Essen. DHEAS was measured by using a competitive chemiluminescence immunoassay (IMMULITE 2000 DHEA-SO4). Estradiol was measured by using either the method of Roche or Siemens (ADVIA Centaur XP) (see Supplementary Material 1). Testosterone was measured using the Siemens method (ADVIA Centaur XP). Hormone levels below the detection limit were replaced (Supplementary Material 1).

Covariates

We considered the following covariates at T2: freezing duration of the blood in month; sex; age in years; body mass index (BMI) calculated from measured height and weight in kg/m²; total years of education according to the International Standard Classification of Education [29], self-reported smoking status (never; former but not within the last year; current smoking within the last year); self-reported sport (yes = one or more sports in the last 4 weeks, else no). The CES-D (Center for Epidemiological Studies Depression Scale) [30] was performed at T2 to detect symptoms of depression. The cut-off value was a score of ≥18 points (depressive symptoms yes, else no). The apolipoprotein E genotype (APOE genotype) with the ɛ2, ɛ3, and ɛ4 alleles was categorized as APOE ɛ4 positive if at least one allele 4 was present (2/4, 3/4, 4/4), else no.

Statistical analysis

The characteristics of the study population are shown in Table 1, presenting the means±standard deviations (SD), medians with quartiles and total numbers of participants with percentages describing the KLK8- and sex hormone distributions and the different covariates. The sex-specific distribution of blood-based KLK8 in CU and CI participants was shown using boxplots. Sex-specific differences of KLK8 were investigated using simple and multiple linear regression to estimate β-coefficients (β) for adjusted mean differences and 95% -confidence intervals ([95%-CI]) adjusted for the experimenter as a proxy for the inter-assay variability, freezing duration of KLK8, depressive symptoms, APOE ɛ4 status, smoking and sport. An additional adjustment for age did not change the effects (data not shown). Scatterplots were performed to show the distribution of the sex hormones DHEAS, estradiol, testosterone and KLK8. The associations between the sex hormones and KLK8 were estimated using simple and multiple linear regression, stratified by sex and adjusted for the covariates inter-assay variability, freezing duration of KLK8, age, BMI, cognitive status, smoking status and sport. The results of the regression analyses are presented in figures showing the estimated β-coefficients (β) and 95% -CI per unit increase of the sex hormones. All reported results of regression analyses refer to the fully adjusted models. The statistical analysis was performed by using R (RStudio, version 1.4.1717), Microsoft Office Excel (version 2205) was used for data processing and Microsoft PowerPoint (version 2311) for data visualization.

Table 1

Characteristics of the study population

Mean±SD; n (%)	Men; n = 159 (54.83)	Women; n = 131 (45.17)	Total; n = 290 (100)
KLK8 [pg/ml]	947.15±795.21	789.37±531.23	875.87±691.99
Median	830.36	827.88	829.12
Q1; Q3	310.42; 1380.13	305.99; 1134.69	305.99; 1263.63
Experimenter
1	69 (43.40)	69 (52.67)	138 (47.59)
2	37 (23.27)	21 (16.03)	58 (20.00)
3	53 (33.33)	41 (31.30)	94 (32.41)
Freezing duration [month]	98.39±10.6	99.39±10.49	98.84±10.54
Age [y]	70.74±6.77	68.52±8.04	69.73±7.44
BMI [kg/m²]	28.59±4.15	27.43±5.08	28.07±4.62
Missing values		2	2
Education [total years]	15.09±2.24	13.63±2.21	14.43±2.34
Smoking status
Never	53 (33.33)	77 (58.78)	130 (44.83)
Former	91 (57.23)	33 (25.19)	124 (42.76)
Current	15 (9.43)	21 (16.03)	36 (12.41)
Sport
Yes	86 (54.01)	75 (57.25)	161 (55.52)
No	73 (45.91)	56 (42.75)	129 (44.48)
Depressive symptoms
No	150 (97.40)	115(91.27)	265 (94.64)
Yes	4 (2.60)	11 (8.73)	15 (5.36)
Missing values	5	5	10
APOEɛ4
Negative	108 (69.68)	91 (72.22)	199 (70.82)
Positive	47 (30.32)	35 (27.78)	82 (29.18)
Missing values	4	5	9
Cognitive status
CU	67 (42.14)	58 (44.27)	125 (43.10)
CI:	92 (57.86)	73 (55.73)	165 (56.90)
– SCD	–45 (48.91)	–28 (38.36)	–73 (44.24)
– aMCI	–23 (25.00)	–14 (19.18)	–37 (22.42)
– naMCI	–24 (26.09)	–31 (42.47)	–55 (33.33)
DHEAS [μg/dl]	104.37±67.65	79.02±51.08	92.97±61.96
Missing values		1	1
Estradiol [pg/ml]	26.01±10.04	13.23±12.63	20.26±12.94
Missing values		1	1
Testosterone [nmol/l]	13.94±5.28	1.41±0.71	8.30±7.39
Missing values		1	1

aMCI, amnestic mild cognitive impairment; APOE ɛ4, apolipoprotein E ɛ4; BMI, body mass index; CI, cognitively impaired; CU, cognitively unimpaired; DHEAS, dehydroepiandrosteronsulfate; IQR, interquartile range; KLK8, kallikrein-8; naMCI, non-amnestic mild cognitive impairment; Q1, 1st quartile; Q3, 3rd quartile; SCD, subjective cognitive decline; SD, standard deviation.

RESULTS

The characteristics of the study population of 290 participants (55% men, 43% CU) are shown in Table 1. Supplementary Table 1 provides the characteristics of the study population stratified by cognitive status. The blood-based KLK8 distribution corresponded to a right skewed distribution (Supplementary Figure 1). The total mean KLK8 concentration±standard deviation (SD) was 875.87±691.99 pg/ml (median: 829.12 pg/ml; Q1 : 305.99 pg/ml; Q3 : 1263.63 pg/ml). The mean KLK8 was lower in women (789.37±531.23 pg/ml) compared to men (947.15±795.21 pg/ml), while the KLK8 medians were similar in women (827.88 pg/ml; Q1 : 305.99 pg/ml; Q3 : 1134.69 pg/ml) and men (830.36 pg/ml; Q1 : 310.42 pg/ml; Q3 : 1380.13 pg/ml). Women had lower sex hormone concentrations of DHEAS (women: 79.02±51.08μg/dl versus men: 104.37±67.65μg/dl), estradiol (13.23±12.63 pg/ml versus 26.01±10.04 pg/ml) and testosterone (1.41±0.71 nmol/l versus 13.94±5.28 nmol/l).

In CU no sex-specific differences in KLK8 could be detected (Fig. 2; women: 795.93±577.71 pg/ml versus men: 808.09±729.55 pg/ml; β [95%-CI]: –95.30 [–324.13; 133.53] pg/ml). However, CI women had on average 262.34 [–493.43; –31.25] pg/ml lower KLK8 values compared to men (Fig. 3a). These results were similar when comparing the CU group with the subgroups SCD, amnestic MCI, and non-amnestic MCI (Supplementary Figures 2 and 3). CI men had on average 220.48 [–25.09; 466.05] pg/ml higher KLK8 values compared to CU men, while in women there was no difference between CU and CI women (36.43 [–165.03; 237.90] pg/ml) (Fig. 3b). The pooled regression model for the total study population also showed that women had lower average KLK8 values compared to men and CI participants had higher KLK8 values compared to CU participants (Supplementary Table 2). This was also shown by an interaction analysis. There seems to be a strong interaction between sex and cognition (Supplementary Table 3). Low precision due to wide confidence intervals with a small sample size should be noted.

Fig. 2

KLK8-distribution in CU versus CI men and women. CI, cognitively impaired; CU, cognitively unimpaired; KLK8, kallikrein-8. Boxplot: big point: mean, crossbar: median, box length: 1st to 3rd quartile, whiskers: 1.5x the box length, diamond: outliers, small points: single values. CU men: n = 67, CU women: n = 58, CI men: n = 92, CI women: n = 73.

Fig. 3

Associations between KLK8 and sex (women versus men), stratified by cognitive status (a) and KLK8 and cognitive status (CI versus CU), stratified by sex (b); results of the multiple linear regression. CI, cognitively impaired; CU, cognitively unimpaired; 95% -CI, 95% -confidence interval; point mark: β-estimate; model 1 (both regressions)): crude; model 2 (3a): fully adjusted for inter-assay variability, freezing duration, depressive symptoms, APOE ɛ4, smoking and sport; model 2 (3b): fully adjusted for inter-assay variability, freezing duration, age, BMI, education, sport, smoking, depressive symptoms and APOE status.

Figure 4 presents the scatterplots of the sex-hormones and KLK8 concentrations. In men there was a weak indication for on average 1.43 [–0.49; 3.34] pg/ml higher KLK8 values per unit increase in DHEAS (Fig. 5a). However, this effect was no longer observed after extreme values (DHEAS > 300μg/dl and KLK8 > 4000 pg/ml) were excluded from the regression analysis (0.30 [–1.70; 2.31] pg/ml, fully adjusted). In men, estradiol was associated with on average 11.94 [–0.42; 24.31] pg/ml higher KLK8 values per unit increase (Fig. 5b). There was no association between testosterone and KLK8 in men (–0.29 [–24.79; 24.20] pg/ml; Fig. 5c) or any associations between sex hormones and KLK8 in women (DHEAS and KLK8: –0.61 [–2.54; 1.33] pg/ml; estradiol and KLK8: –1.66 [–9.30; 5.98] pg/ml; testosterone and KLK8: –36.35 [–169.31; 97.12] pg/ml; Fig. 5a–c). Stratified by cognitive subgroups, the association between DHEAS and KLK8 in men seemed to occur mainly in CU and the association between estradiol and KLK8 in men was strongest in the non-amnestic MCI group (Supplementary Figures 4–6).

Fig. 4

Scatterplots of the hormone and KLK8 concentrations in men (on the left side) versus women (on the right side). DHEAS, dehydroepiandrosteronsulfate; KLK8, kallikrein-8. Drawn line: regression line of unadjusted simple linear regression. men: n = 159, women: n = 130.

Fig. 5

Association between KLK8 and the sex hormones DHEAS (5a), estradiol (5b) and testosterone (5c), stratified by sex; results of the multiple linear regression. KLK8, kallikrein-8; DHEAS, dehydroepiandrosteronsulfate; 95% -CI, 95% -confidence interval; point mark: β-estimate; model 1: crude; model 2: fully adjusted for inter-assay variability, freezing duration, age, BMI, cognitive status, smoking, and sport.

DISCUSSION

Sex-specific blood-based KLK8 differences in CU and CI participants

There were no sex-specific blood-based KLK8 differences in CU. These results were similar to those of the studies by Teuber-Hanselmann and Goldhardt mentioned above, which have not found sex-differences in cerebrospinal fluid and blood-based KLK8 in CU but also in patients with MCI or AD [2, 10]. This contradicts results of other studies, which showed higher KLK8 concentrations in women's neuronal tissues and in female mice compared to male [3, 9]. Since we only measured blood-based KLK8 (which is the sum of the total expression of all KLK8-expressing tissues [31]) the results of our study can be compared with previous studies only to a limited extent.

In CI, women had lower average KLK8 values compared to men. This finding contradicts the prior suggestion that women have higher KLK8 values than men, due to higher estradiol concentrations [3, 9]. In our study CI men had higher average KLK8 levels compared to CU men, while in CU and CI women there were no differences. This was also shown by the interaction analysis. Since there are biological differences in the brains of men and women in anatomical, psychological, behavioral, and cognitive aspects [11] and since AD itself shows sex-specific differences like the progression rates from MCI to AD [11], the atrophy of the hippocampus [12] or the disease survival [11], we assume that KLK8 may play a different role in the pathology of AD in men and women.

DHEAS and KLK8

In our study, there was weak evidence for a positive slope between DHEAS and blood-based KLK8 in men. DHEAS serves primarily as a precursor for other sex hormones [32], has neurotrophic effects [33] and an age-related decrease in DHEAS is associated with increasing cognitive impairment [34]. Thus, a connection to KLK8 and cognition is conceivable; but a mechanism in which DHEAS and KLK8 interact is unknown so far. The apparent anti-dementive effects of DHEAS [34] contradict our results, showing a positive slope with the possibly AD-promoting KLK8. In women, no associations between DHEAS and KLK8 could be found, which could be explained by the fact that old women use to have lower DHEAS-Levels compared to men [35], like in our study, and those low DHEAS-levels may not have been able to show a statistically detectable effect on KLK8.

Estradiol and KLK8

After some demonstrations that estradiol has an increasing influence on KLK8 in cell cultures [9 , 36], a positive slope between estradiol and KLK8 was also found in this analysis, at least in men. While this finding in the male study population is consistent with the present theory of estradiol being associated with higher KLK8 levels [9 , 36], there was no association between estradiol and KLK8 in the female study population. A possible explanation may be the low and homogeneously distributed concentrations of estradiol in women due to the postmenopause status of the women.

The results agree only in part with the prior assumption that estradiol increases KLK8 with a following higher risk for AD [3]. But, estradiol itself is assumed to be neuroprotective [21] and lower estrogen levels (as a result of menopause) are associated with a higher risk for cognitive impairment and AD [37]. Especially the excessive decrease of estradiol in women within aging, compared to a slowly decreasing testosterone level in men [11, 38], is expected to lead to a shift in the brain metabolism, to lead to oxidative stress and hypometabolism, to damage the brain cells and to lead to cognitive impairment [11, 39]. Controversial evidence regarding hormone replacement therapies for neuroprotection led to the “window theory”, which proposes that middle age around menopause is the critical period for the development of dementia and that estrogen therapy may only protect against the increased risk of cognitive impairment during this early period [11 , 39]. This could be a possible explanation for the contrast between the neuroprotective role and possibly AD-promoting KLK8-overexpression by estradiol. It is hypothesized that hormones act differently in the brains of older compared to younger persons due to altered receptor distribution and density. An initially neuroprotective function of estradiol may no longer be effective, or may even harm brain cells, due to decreased tissue sensitivity with age [37]. As we have a study population with old participants with generally lower sex hormone levels, it is therefore possible that estrogen has increased KLK8 in men because of the AD pathophysiology and that women have lower KLK8 values because of lower estrogen levels. In addition, three women of the study population who were taking estrogens at time T2 had not only higher mean estradiol (51.27 pg/ml versus 12.33 pg/ml) but also mean KLK8 levels (1117.84 pg/ml versus 781.67 pg/ml) compared to women without estrogen intake (Supplementary Table 4), also suggesting that a positive association between estradiol and KLK8 exists but could not be demonstrated in or female study population because of low estradiol levels.

Testosterone and KLK8

We could not detect any associations between testosterone and KLK8 in men or women in or study. The result is similar to the experiments of Keyvani et al. in which testosterone failed to increase KLK8 levels in neuronal and microglial cell cultures [9], but it contrasts with the demonstrated increase of blood KLK8 by testosterone treatment of female-to-male transsexuals [16]. However, since the in vivo study by Slagter et al. involved an artificial supply of testosterone [16] the results cannot directly be compared with the “natural” hormone concentrations of the participants in our study. We cannot exclude an effect of testosterone on KLK8 at any age, because of the advanced age of our study population and resulting lower testosterone levels.

Sex-specific differences in blood-based KLK8, influenced by sex hormones

The findings of our exploratory study support the assumption of KLK8 being influenced by sex hormones. The hormones could be considered as the cause for a higher KLK8-concentration in the male study population, as those had the higher average concentrations of all three hormones, compared to women. However, the KLK8 concentration of the male compared to the female population was only higher in CI, which is why the hormones do not provide a sufficient explanation for the gender difference alone. Conceivably, hormones could have a differential effect in a healthy compared to a pathologically damaged brain tissue, similar to the theory that the protective direct effects of hormones in the CNS could lead to damage with age, or that the sensitivity of the tissue changes with increasing pathology [37]. Assuming that the association between KLK8 and hormones also underlies such effects, the interaction between KLK8 and the hormones may be differed in CU and CI.

Strength and limitations

This exploratory study is the first one investigating sex-specific differences in blood-based KLK8 by investigating KLK8 and sex hormones in participants of a population-based study with different cognitive status. Since findings in mouse models are only limitedly transferrable to humans [40], this analysis provides important additional information. But it should be noticed that the measurement of KLK8 was originally performed for another analysis and there were many excluding criteria (see “study population”), so the analysis does not provide the general population. Previous studies are referring to an association between KLK8-overexpression and AD, while in our study, only participants with possible AD precursors are represented. Our cognitive diagnoses were based on cognitive tests only. We did not have any AD biomarker information. Thus, we cannot state that our participants in the CI group were on the AD's continuum. Furthermore, participants with SCD were included, although it is unknown if they ever show a progression into MCI or dementia. The accuracy of the KLK8 measurements could be affected by the freezing duration of the serum with the possibility of falsification by pre-analytical treatment [41]. In our female study population, the estradiol levels were very low, which can be explained by the fact that 86% of the women were postmenopausal; we could not investigate possible earlier effects. Finally, the distribution of KLK8 was right-skewed and it was shown that the association between DHEAS and estradiol and KLK8 in men were influenced by outliers. But, when dividing the study population into KLK8 quartiles according to sex, it can also be seen that the mean values of DHEAS and estradiol in men increased with increasing quartiles (Supplementary Table 5). Logarithmic transformation of KLK8 did not provide a better model fit (data not shown). In consideration of those results, all interpretations of this exploratory study were formulated carefully with the awareness that some associations cannot be detected in the transformed models and are therefore uncertain.

Conclusion

Because CI women had lower mean blood-based KLK8 levels compared to CI men and because of existing sex differences in the pathology of dementia [11], KLK8 may play a different role in the development of cognitive impairment in men and women. The reasons for these sex-differences might be due to the concentration of steroid hormones, affecting KLK8 levels in blood. The results of our exploratory study suggest that DHEAS and estradiol might influence the KLK8 level in men. However, further research on hormonal regulation of KLK8 expression and other influencing factors is needed to use blood-based KLK8 as an early AD biomarker and to properly interpret KLK8 blood levels. Therefore, future studies may consider sex-specific serum-KLK8 differences as a part of the investigation of the KLK8 involvement in cognitive impairment and AD pathology.

AUTHOR CONTRIBUTIONS

Nela Krizanovic (Conceptualization; Data curation; Formal analysis; Methodology; Validation; Visualization; Writing – original draft); Martha Jokisch (Writing – review & editing); Karl-Heinz Jöckel (Conceptualization; Methodology; Writing – review & editing); Börge Schmidt (Methodology; Writing – review & editing); Andreas Stang (Methodology; Writing – review & editing); Sara Schramm (Conceptualization; Methodology; Project administration; Supervision; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

Our thanks go to all participants of the HNR study, the study founders, financial supporters, and the HNR personnel at the study center in Essen. We thank Prof. Kathy Keyvani and the personnel of the laboratory of the Institute of Neuropathology of the University Hospital Essen for the measurement of the KLK8. The authors also thank the Advisory Board of the HNR Study: T. Meinertz, Hamburg, Germany (Chair); C. Bode, Freiburg, Germany; P.J. de Feyter, Rotterdam, Netherlands; B. Güntert, Hall i.T., Austria; F. Gutzwiller, Bern, Switzerland; H. Heinen, Bonn, Germany; O. Hess (†), Bern, Switzerland; B. Klein (†), Essen, Germany; H. Löwel, Neuherberg, Germany; M. Reiser, Munich, Germany; G. Schmidt (†), Essen, Germany; M. Schwaiger, Munich, Germany; C. Steinmüller, Bonn, Germany; T. Theorell, Stockholm, Sweden; and S.N Willich, Berlin, Germany.

This manuscript was created as a part of the doctoral thesis of Nela Krizanovic.

FUNDING

The authors thank the Heinz Nixdorf Foundation [Chairman: Martin Nixdorf; Past Chairman: Dr jur. Gerhard Schmidt (†)], for their generous support of this study. Parts of the study were also supported by the German Research Council (DFG) [DFG project: EI 969/2-3, ER 155/6-1;6-2, HO 3314/2-1;2-2;2-3;4-3, INST 58219/32-1, JO 170/8-1, KN 885/3-1, PE 2309/2-1, SI 236/8-1;9-1;10-1,], the German Ministry of Education and Science [BMBF project: 01EG0401, 01GI0856, 01GI0860, 01GS0820_WB2-C, 01ER1001D, 01GI0205], the Ministry of Innovation, Science, Research and Technology, North Rhine-Westphalia (MIWFT-NRW), the Else Kröner-Fresenius-Stiftung [project: 2015_A119], the German Social Accident Insurance [DGUV project: FF-FP295], and the German Heart Fondation. Furthermore the study was supported by the Competence Network for HIV/AIDS, the deanship of the University Hospital and IFORES of the University Duisburg-Essen, the European Union, the German Competence Network Heart Failure, Kulturstiftung Essen, the Protein Research Unit within Europe (PURE), the Dr. Werner-Jackstädt Stiftung and the following companies: Celgene GmbH München, Imatron/GE-Imatron, Janssen, Merck KG, Philips, ResMed Foundation, Roche Diagnostics, Sarstedt AG&Co, Siemens HealthCare Diagnostics, Volkswagen Foundation.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The corresponding author has full access to all data in the study. Due to data security reasons the HNR study does not allow sharing data as a public use file. Data requests can be addressed to recall@uk-essen.de.

The supplementary material is available in the electronic version of this article: .

References

Mella

, Figueroa

, Otth

, Ehrenfeld

(2020) Involvement of kallikrein-related peptidases in nervous system disorders. Front Cell Neurosci 14, 166.

Teuber-Hanselmann

, Rekowski

, Vogelgsang

, von Arnim

, Reetz

, Stang

, Jöckel

, Wiltfang

, Esselmann

, Otto

, Tumani

, Herring

, Keyvani

(2020) CSF and blood kallikrein-8: A promising early biomarker for Alzheimer’s disease. J Neurol Neurosurg Psychiatry 91, 40–48.

Herring

, Kurapati

, Krebs

, Grammon

, Scholz

, Voss

, Miah

, Budny

, Mairinger

, Haase

, Teuber-Hanselmann

, Dobersalske

, Schramm

, Jöckel

, Münster

, Keyvani

(2021) Genetic knockdown of Klk8 has sex-specific multi-targeted therapeutic effects on Alzheimer’s pathology in mice. Neuropathol Appl Neurobiol 47, 611–624.

Schramm

, Jokisch

, Jöckel

, Herring

, Keyvani

(2021) Is kallikrein-8 a blood biomarker for detecting amnestic mild cognitive impairment? Results of the population-based Heinz Nixdorf Recall study. Alzheimers Res Ther 13, 202.

Jessen

, Amariglio

, Buckley

, van der Flier

, Han

, Molinuevo

, Rabin

, Rentz

, Rodriguez-Gomez

, Saykin

, Sikkes

SAM

, Smart

, Wolfsgruber

, Wagner

(2020) The characterisation of subjective cognitive decline. The Lancet Neurology 19, 271–278.

Jokisch

, Krizanovic

, Jöckel

K-H

, Herring

, Keyvani

, Weimar

, Schramm

, Group tHNRSI (2022) Is blood-based kallikrein-8 increased in participants with subjective cognitive decline compared to cognitively healthy controls? Results of the Heinz Nixdorf Recall Study. Alzheimers Dement 18, e059955.

Roberts

, Knopman

(2013) Classification and epidemiology of MCI. Clin Geriatr Med 29, 753–772.

Schramm

, Krizanovic

, Roggenbuck

, Jöckel

, Herring

, Keyvani

, Jokisch

(2023) Blood kallikrein-8 and non-amnestic mild cognitive impairment: An exploratory study. J Alzheimers Dis Rep 7, 327–337.

Keyvani

, Münster

, Kurapati

, Rubach

, Schönborn

, Kocakavuk

, Karout

, Hammesfahr

, Wang

, Hermann

, Teuber-Hanselmann

, Herring

(2018) Higher levels of kallikrein-8 in female brain may increase the risk for Alzheimer’s disease. Brain Pathol 28, 947–964.

10.

Goldhardt

, Warnhoff

, Yakushev

, Begcevic

, Förstl

, Magdolen

, Soosaipillai

, Diamandis

, Alexopoulos

, Grimmer

(2019) Kallikrein-related peptidases 6 and 10 are elevated in cerebrospinal fluid of patients with Alzheimer’s disease and associated with CSF-TAU and FDG-PET. Transl Neurodegener 8, 25.

11.

Snyder

, Asthana

, Bain

, Brinton

, Craft

, Dubal

, Espeland

, Gatz

, Mielke

, Raber

, Rapp

, Yaffe

, Carrillo

(2016) Sex biology contributions to vulnerability to Alzheimer’s disease: A think tank convened by the Women’s Alzheimer’s Research Initiative. Alzheimers Dement 12, 1186–1196.

12.

Zhu

, Montagne

, Zhao

(2021) Alzheimer’s pathogenic mechanisms and underlying sex difference. Cell Mol Life Sci 78, 4907–4920.

13.

Shaw

, Diamandis

(2008) Regulation of human tissue kallikrein-related peptidase expression by steroid hormones in 32 cell lines. Biol Chem 389, 1409–1419.

14.

Paliouras

, Diamandis

(2007) Coordinated steroid hormone-dependent and independent expression of multiple kallikreins in breast cancer cell lines. Breast Cancer Res Treat 102, 7–18.

15.

Kishi

, Grass

, Soosaipillai

, Scorilas

, Harbeck

, Schmalfeldt

, Dorn

, Mysliwiec

, Schmitt

, Diamandis

(2003) Human kallikrein 8, a novel biomarker for ovarian carcinoma. Cancer Res 63, 2771–2774.

16.

Slagter

, Scorilas

, Gooren

, de Ronde

, Soosaipillai

, Giltay

, Paliouras

, Diamandis

(2006) Effect of testosterone administration on serum and urine kallikrein concentrations in female-to-male transsexuals. Clin Chem 52, 1546–1551.

17.

Yeung

CHC

, Au Yeung

, Kwok

, Zhao

, Schooling

(2023) The influence of growth and sex hormones on risk of alzheimer’s disease: A mendelian randomization study. Eur J Epidemiol 38, 745–755.

18.

Gouras

, Xu

, Gross

, Greenfield

, Hai

, Wang

, Greengard

(2000) Testosterone reduces neuronal secretion of Alzheimer’s beta-amyloid peptides. Proc Natl Acad Sci U S A 97, 1202–1205.

19.

Bianchi

(2022) Impact of testosterone on Alzheimer’s disease. World J Mens Health 40, 243–256.

20.

Pike

, Nguyen

, Ramsden

, Yao

, Murphy

, Rosario

(2008) Androgen cell signaling pathways involved in neuroprotective actions. Horm Behav 53, 693–705.

21.

McEwen

, Milner

(2017) Understanding the broad influence of sex hormones and sex differences in the brain. J Neurosci Res 95, 24–39.

22.

Schmermund

, Möhlenkamp

, Stang

, Grönemeyer

, Seibel

, Hirche

, Mann

, Siffert

, Lauterbach

, Siegrist

, Jöckel

K-H

, Erbel

(2002) Assessment of clinically silent atherosclerotic disease and established and novel risk factors for predicting myocardial infarction and cardiac death in healthy middle-aged subjects: Rationale and design of the Heinz Nixdorf RECALL Study. American Heart Journal 144, 212–218.

23.

Stang

, Moebus

, Dragano

, Beck

, Möhlenkamp

, Schmermund

, Siegrist

, Erbel

, Jöckel

(2005) Baseline recruitment and analyses of nonresponse of the Heinz Nixdorf Recall Study: Identifiability of phone numbers as the major determinant of response. Eur J Epidemiol 20, 489–496.

24.

Müller-Gerards

, Weimar

, Abramowski

, Tebrügge

, Jokisch

, Dragano

, Erbel

, Jöckel

, Moebus

, Winkler

(2019) Subjective cognitive decline, APOE ɛ4, and incident mild cognitive impairment in men and women. Alzheimers Dement (Amst) 11, 221–230.

25.

Wege

, Dlugaj

, Siegrist

, Dragano

, Erbel

, Jöckel

, Moebus

, Weimar

(2011) Population-based distribution and psychometric properties of a short cognitive performance measure in the population-based Heinz Nixdorf Recall Study. Neuroepidemiology 37, 13–20.

26.

Association AP (1994) Diagnostic and statistical manual of mental disorders (4th ed.) American Psychiatric Publishing, Inc.

27.

(2011) The selection and use of essential medicines. World Health Organ Tech Rep Ser i-xiv, 1–249.

28.

Winblad

, Palmer

, Kivipelto

, Jelic

, Fratiglioni

, Wahlund

, Nordberg

, Bäckman

, Albert

, Almkvist

, Arai

, Basun

, Blennow

, de Leon

, DeCarli

, Erkinjuntti

, Giacobini

, Graff

, Hardy

, Jack

, Jorm

, Ritchie

, van Duijn

, Visser

, Petersen

(2004) Mild cognitive impairment–beyond controversies, towards a consensus: Report of the International Working Group on Mild Cognitive Impairment. J Intern Med 256, 240–246.

29.

Statistics, UNESCO (1997) International Standard Classification of Education (ISCED). UNESCO, Montreal, QC, Canada.

30.

Radloff

(1977) The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Measure 1, 385–401.

31.

Kishi

, Grass

, Soosaipillai

, Shimizu-Okabe

, Diamandis

(2003) Human kallikrein 8: Immunoassay development and identification in tissue extracts and biological fluids. Clin Chem 49, 87–96.

32.

van den Beld

, Kaufman

, Zillikens

, Lamberts

SWJ

, Egan

, van der Lely

(2018) The physiology of endocrine systems with ageing. Lancet Diabetes Endocrinol 6, 647–658.

33.

Baulieu

E-E

, Robel

(1998) Dehydroepiandrosterone (DHEA) and dehydroepiandrosterone sulfate (DHEAS) as neuroactive neurosteroids. Proc Natl Acad Sci U S A 95, 4089–4091.

34.

Pan

, Wu

, Kaminga

, Wen

, Liu

(2019) Dehydroepiandrosterone and dehydroepiandrosterone sulfate in Alzheimer’s disease: A systematic review and meta-analysis. Front Aging Neurosci 11, 61.

35.

Chahal

, Drake

(2007) The endocrine system and ageing. J Pathol 211, 173–180.

36.

Yousef

, Diamandis

(2001) The new human tissue kallikrein gene family: Structure, function, and association to disease. Endocr Rev 22, 184–204.

37.

Barron

, Pike

(2012) Sex hormones, aging, and Alzheimer’s disease. Front Biosci (Elite Ed) 4, 976–997.

38.

Wang

, Chen

, Ye

, Zirkin

, Chen

(2017) Steroidogenesis in Leydig cells: Effects of aging and environmental factors. Reproduction 154, R111–R122.

39.

Morgan

, Derby

, Gleason

(2018) Cognitive changes with reproductive aging, perimenopause, and menopause. Obstet Gynecol Clin North Am 45, 751–763.

40.

Dubal

, Broestl

, Worden

(2012) Sex and gonadal hormones in mouse models of Alzheimer’s disease: What is relevant to the human condition? Biol Sex Differ 3, 24.

41.

Kisand

, Kerna

, Kumm

, Jonsson

, Tamm

(2011) Impact of cryopreservation on serum concentration of matrix metalloproteinases (MMP)-7, TIMP-1, vascular growth factors (VEGF) and VEGF-R2 in Biobank samples. Clin Chem Lab Med 49, 229–235.